E.A. Schneidmiller and M.V. Yurkov

FEL R&D Meeting, October 14, 2020

Options for generation of attosecond

x-ray pulses at the European XFEL

Why we need short radiation pulses?

Radiation sources with ultrashort pulse duration (down to attosecond time scale) are

required for studying dynamical process in the matter:

• Phase transitions in solids.

• Surface dynamics.

• Making and braking of bonds during chemical reactions.

• Chemical dynamics in proteins.

• Correlated behavior of electrons in complex solids .

…. And many others.

Temporal scale depends on physical phenomena under investigation. Typical atomic

time scale is in a few tens of attosecond range; vibration period between the two

hydrogen atoms in a hydrogen molecule is about 8 fs; 1Å / (speed of sound) ~ 100 fs.

Typical experiment for studying time-dependent phenomena involves two radiation

pulses (pump-probe technique) when the first pulse triggers some physical process,

and the second pulse is used for probing state of the system at prescribed moment of

time. Pulse duration and synchronization accuracy are key parameters for achieving

temporal resolution.

Requirements to the radiation source

Scope of requirements to the radiation source:

• Photon energy range;

• Temporal, spectral and polarization properties;

• Intensity (photons per pulse);

• Repetition rate;

• Synchronization with an external laser or other x-ray pulse (pump-probe

experiments);

• Stability (intensity, position, temporal);

• Contrast of an ultrashort (e.g., attosecond) pulse wrt background (radiation

pedestal);

• Photon beam transport to a sample;

• Need in a monochromator (intensity reduction and radiation damage).

J. Levesque and P.B. Corkum, Canadian Journal of Physsics, 84(2006)1

A. Zholents, Proc. FEL2010 Conf.

2010

Thomson

Slicing

SPPS

LCLS (FEL)

x-rays

(from electron beams)Visible

(lasers)

FLASH (FEL)

VUV/soft x-rays

(HHG)

Status of ultrashort pulses production

Production of ultrashort radiation pulses by

electron beams

• Production of ultrashort electron pulses which then generate ultrashort

radiation pulses.

• Slicing of electron bunches aiming local change of electron beam properties

at a short time scale.

• Then different techniques (based on physical effects as well as technical

tricks) are used to derive ultrashort radiation pulse.

• Coherent enhancement of the radiation intensity is an essential issue.

6

Some examples of ultrashort pulse

generation

• “Simple” schemes (not involving external lasers)

- Statistical selection

- Slotted foil

- Low charge

- LCLS experience (nonlinear compression and XLEAP)

• (Some of the) schemes using few-cycle laser pulse

- Chirp and taper

- Current enhancement

Formation of ultrashort bunches in linear

accelerators (x-ray FELs, linear colliders)

• Application of several stages of bunch compression. Electron beam

gains energy chirp in acceleration section which then is transformed

into spatial compression in magnetic chicane.

• Strongly nonlinear compression can produce very sharp features of

the electron beam /W. Ackermann et al, Nature Photonics

1(2007)336/.

• Linearization of compression scheme with higher harmonics rf

cavities allows to compress significant fraction of the electron bunch.

Tail particle, more momentum

Head particle, less momentum

FIR

To pump-probe

experiment

8

Direct production (low charge)

S. Reiche et al., NIM A593(2008)45

1 pC case simulated

LCLS operates at 20 pC, few fs x-rays

Y. Ding et al., PRL 102(2009)254801

Talk by Z. Huang

9

Slotted foil method

PRL 92, 074801 (2004)P. Emma et al.,

(femtoseconds)

P. Emma, Z. Huang, M. Borland, FEL’2004

(attoseconds)

10

Methods using few-cycle laser pulse

Laser pulse with carrier-envelope

phase (CEP) stabilization (~ 1 mJ)Energy modulation of electron beam

A. Zholents and W. Fawley, PRL 92(2004)224801

E. Saldin, E. Schneidmiller and M. Yurkov, Opt. Comm. 237(2004)153

E. Saldin, E. Schneidmiller and M. Yurkov, Opt. Comm. 239(2004)161

A. Zholents and G. Penn, PRST-AB 8(2005)050704

E. Saldin, E. Schneidmiller and M. Yurkov, PRST-AB 9(2006)050702

A. Zholents and M. Zolotorev, New J. Phys. 10(2008)025005

Y. Ding et al., PRST-AB 12(2009)060703

D. Xiang, Z. Huang and G. Stupakov, PRST-AB 12(2009)050702

Two-period undulator

11

Energy chirp and undulator taper

E. Saldin, E. Schneidmiller and M. Yurkov, PRST-AB 9(2006)050702

High power, high

contrast attosecond

pulses

Energy chirp is perfectly compensated for

by undulator taper

12

Current enhancement

A. Zholents , PRST-AB 8(2005)040701; A. Zholents et al., FEL2004

Take care of LSC in the

undulator! (G.Geloni et

al., NIM A583(2007)228)

13

Recent LCLS experiments

• Nonlinear compression + LSC + undulator taper: worked in

hard X-ray regime

S. Huang et al., Phys. Rev.Lett.119(2017)154801

• XLEAP (“XCHEAP”): self-modulation of beam in the wiggler +

compression of a short slice + LSC + taper: sucsessfully worked in

soft X-ray regime

J. Duris et al.,

Nat. Photon.1430(2020)6

14

Statistical selection of a single spike

8 Å

I8 I18

1 Å

300 as

E. Saldin, E. Schneidmiller, M. Yurkov, Opt. Comm. 212(2002)377

15

Options for EuXFEL

• Nonlinear compression and/or low charge (~ 1 fs): studies by Bolko and

colleagues.

• The same but in combination with harmonic cascade/statistical selection

in SASE3: attosecond pulses. For example: 700 eV – 2.1 keV – 4.2 keV.

Can be tried out in the near future.

• Slotted foil is of limited applicability at the European XFEL – does not

survive at high repetition rate (high average power).

• XLEAP-like approach (with a wiggler for self-modulation) can be studied

but we need to accumulate a culture of bunch shape manipulation (stable

and reproducible horn)

• Laser-based attosecond schemes should be seriously considered.

Synchronization on 10 fs scale is state-of-the-art (combination with slotted

foil a la LCLS is not required). Modulator undulator and chicane are

simple and inexpensive but the laser transport and infrastructure require a

lot of efforts.